Laminin and β1 Integrins Are Crucial for Normal Mammary Gland Development in the Mouse

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Laminin and β1 Integrins Are Crucial for Normal Mammary Gland Development in the Mouse
  Lamininand  1IntegrinsAreCrucial forNormalMammaryGlandDevelopmentintheMouse TeresaC.M.Klinowska, 1  JesusV.Soriano,* GwynnethM.Edwards, JanineM.Oliver,AnthonyJ.Valentijn,RobertoMontesano,*andCharlesH.Streuli School of Biological Sciences, U niversity of Manchester, 3.239 Stopford Building, Oxford Road,Manchester, M13 9PT, U nited Kingdom; and   * Department of Morphology, U niversity Medical Centre, 1 rue Michel-Servet, 1211 Geneva 4, Switzerland  Wehaveexaminedtheroleofintegrin–extracellularmatrixinteractionsinthemorphogenesisofductal structures in vivo  usingthedevelopingmousemammaryglandasamodel.Atpuberty,ductal growthfromterminal endbudsresultsinanarborescentnetworkthateventuallyfillsthegland,whereuponthebudsshrinkinsizeandbecomemitoticallyinactive.Endbudsaresurroundedbyabasementmembrane,whichweshowcontainslaminin-1andcollagenIV.Toaddresstheroleof cell–matrixinteractionsinglanddevelopment,pelletscontainingfunction-perturbinganti-  1integrin,anti-  6integrin,andanti-lamininantibodiesrespectivelywereimplantedintomammaryglandsatpuberty.Blocking  1integrinsdramaticallyreducedboththenumberofendbudsperglandandtheextentofthemammaryductal network,comparedwithcontrols.Theseeffectswerespecifictotheendbudssincetherestoftheglandarchitectureremainedintact.Reduceddevelopmentwasstill apparent after6days,but endbudssubsequentlyreappeared,indicatingthat theinhibitionof   1integrinswasreversible.Similarresultswereobtainedwithanti-lamininantibodies.Incontrast,noeffectonmorphogenesis in vivo   wasseenwithanti-  6integrinantibody,suggestingthat  6isnottheimportantpartnerfor  1inthissystem.Thestudieswith  1integrin wereconfirmedin aculturemodel of ductal morphogenesis, whereweshow that hepatocytegrowth factor(HGF)–inducedtubulogenesisisdependentonfunctional  1integrins.ThusintegrinsandHGFcooperatetoregulateductalmorphogenesis.Weproposethatbothlamininand  1integrinsarerequiredtopermitcellulartractionthroughthestromalmatrix andarethereforeessential for maintainingendbudstructureandfunction in normal pubertal mammary glanddevelopment.  ©1999AcademicPress K ey Words:   integrin;mousemammarygland;morphogenesis;hepatocytegrowthfactor;extracellularmatrix;laminin. INTRODUCTION Tissue morphogenesis relies on the dynamic interactionsof cells w ith each other and w ith the extracellular matrix(ECM). 2 Key molecules involved in these interactions arethe integrin family of cell surface ECM receptors. Mostintegrin functions have been defi ned using culture modelsand blocking antibodies, disintegrins, antisense constructs,or peptides corresponding to binding sites (Adler and Chen,1992; Lallier  et al.,  1992; Trikha  et al.,  1994). Such inter-ference w ith cell–ECM communication has been demon-strated to affect cell adhesion, migration, differentiationstate, gene expression, proliferation, and survival (review edin Ashkenas  et al.,  1996). How ever, it is still largelyunresolved as to w hether integrins are involved in similarmechanisms  in vivo. A number of studies have explored the effect of injectingintegrin-blocking antibodies intravenously. These haveconcentrated on the subsequent behaviour of cells of theimmune system (Abraham  et al.,  1994; G orcyznski  et al., 1995; Weg  et al.,  1993) or the degree of metastatic spreadfrom primary tumours (Elliott  et al.,  1994), w hich areessentially effects on single cells or small groups of cells.How ever, most cells  in vivo   do not exist singly but rather as 1 To w hom correspondence should be addressed. Fax:   44 161275 3915. E-mail: tklinow 2 Abbreviations used: ECM, extracellular matrix; EHS, En-gelbreth–Holm–Sw arm; H&E, haematoxylin and eosin; HG F, he-patocyte grow th factor; TG F, transforming grow th factor; SDS,sodium dodecyl sulphate; PAG E, polyacrylamide gel electrophore-sis. Developmental Biology  215, 13–32 (1999)Article ID dbio.1999.9435, available online at http://w w w on 0012-1606/99 $30.00Copyright © 1999 by Academic PressAll rights of reproduction in any form reserved.  13  a three-dimensional multicellular architecture, and few studies have explored how integrin function relates to thedevelopment and maintenance of tissue structure.To investigate the importance of integrins in tissuemorphogenesis, their role in developing systems  in vivo  needs to be examined and to this end, knockout mice havebeen created by homologous recombination. However,many homozygous integrin deletions are lethal at earlystages of development (reviewed by Hynes, 1996, andFassler  et al.,  1996). Thus in order to study the function ofintegrins in late-developing organ systems such as themammary gland, other approaches are required. The injec-tion of function-blocking antibodies into specifi c sites ofinterest has been used in several developmental models. Innonmammalian systems, the presence of anti-  1 integrinantibody blocks interstitial cell migration of  H ydra   grafts  in vivo   (Agbas and Sarras, 1994), while injection of anti-  1antibodies into the blastocoel of amphibian embryos dis-rupts gastrulation and mesodermal cell migration (Boucaut et al.,  1984; Howard  et al.,  1992; Johnson  et al.,  1993). Inavian embryos, introduction of antibodies to   1 integrinsinto cranial neural crest cell migratory pathways results inserious perturbation of normal migration (Bronner Fraser,1985). However, this approach has not previously been usedto investigate integrin function in developing mammaliantissues  in vivo. In this study we have used the mammary gland as amodel system to examine the role of integrin–matrix inter-actions in ductal morphogenesis. At birth the murinemammary gland is a simple rudiment consisting of a few ducts that extend from the nipple a short distance into thesubcutaneous fat pad. G landular growth then arrests untilthe onset of puberty at around 3 weeks of age when cells inbulbous structures at the ductal tips, known as end buds,begin to proliferate, resulting in lengthening and branchingof the epithelial ductal network. This growth results in anarborescent network of interconnecting tubes, which even-tually fi lls the subcutaneous fatty stroma of the gland. Inareas where there is no room for further ductal expansion,the end buds regress to become quiescent terminal struc-tures. By contrast, active terminal end buds are highlymitotic, multilayered epithelial structures found only dur-ing ductal growth. They consist of a teardrop-shaped groupof cells containing 6–10 layers of body or luminal cells.This is surrounded by an outer layer of cap cells at the tip ofthe end bud and myoepithelial cells at its neck and extend-ing along the length of the duct. These lie on a specialisedform of ECM, the basement membrane. The duct in turn isensheathed in a collagenous stroma that extends along itslength up to the neck of the end bud. The fi brous stromamay constrict the direction of growth by only allowingductal elongation from the tip of the end bud (Williams andDaniel, 1983); side branches form in areas of breaks in thesheath. The mammary ductal network is in turn sur-rounded by the adipose cells of the fat pad.A number of factors have been shown to affect end budformation and ductal elongation including growth factorssuch as transforming growth factor (TG F)    and   , epider-mal growth factor, and hepatocyte growth factor (HG F),hormones such as oestrogen and growth hormone, and celladhesion molecules including E- and P-cadherin (Coleman et al.,  1988; Daniel  et al.,  1995; Kleinberg, 1997; Silberstein and Daniel, 1987; Snedeker  et al.,  1991; Yang  et al.,  1995).We now present a study examining the role of integrin–matrix interactions in the developing virgin mammarygland  in vivo.  Earlier work using culture models of primarymammary epithelial cells isolated from pregnant mice hasdemonstrated a role for integrin–ECM interactions in thesuppression of apoptosis and also in milk production at alater stage of mammary gland development (Pullan  et al., 1996; Streuli  et al.,  1991), and   1 integrin has been shownto be required for full pregnancy–associated development  in vivo   (Faraldo  et al.,  1998). However, it has not previouslybeen determined whether integrin receptors have an addi-tional role in the control of ductal elongation during devel-opment in the nonpregnant pubertal virgin animal.To address this question, we examined the effect offunction-blocking anti-  1 and anti-  6 integrin antibodieson ductal morphogenesis and report here for the fi rst timethat blocking   1 but not   6 integrin dramatically reducesthe number of end buds in the gland  in vivo   compared withcontrols. Furthermore, we show that the anti-  1 integrinantibody disrupts morphogenesis in a culture model ofmammary tubulogenesis. Similar results were also obtained in vivo   with a function-blocking anti-laminin antibody.These effects were not due to simple disruption of allcell–matrix interactions since the mammary gland retainedits histoarchitecture; the effects observed were specifi c tothe developing buds. We therefore propose that contactwith the extracellular matrix, and with laminin in particu-lar, via   1 integrins is a key factor in maintaining end budstructure and function, and thus that normal developmentand ductal morphogenesis of the mammary gland dependon functional   1 integrins. We also demonstrate that thekey partner for these interactions is not the   6 integrinsubunit. MATERIALS AND METHODS Antibody Preparation  The detailed method for the preparation of the anti-  1 integrinantibody is published elsewhere (Edwards and Streuli, 1999). Inbrief, rat monoclonal antibody PS/2 (anti-  4 integrin) was used topurify   4  1 integrin from E14.5–16.5 mouse embryo homogenate.After prebleeding, two New Zealand White rabbits were immu-nized with 50   g purifi ed   4  1 integrin and boosted after 12 weeks.Rabbit IgG was purifi ed by protein A chromatography, but notfurther purifi ed on an integrin column because we wished to retainas much of the integrin function-perturbing component of theantiserum as possible. An antibody to the COOH-terminal domainof the mouse   1 integrin subunit was developed in rabbits andaffi nity-purifi ed on an integrin peptide column as described previ-ously (Delcommenne and Streuli, 1995). Rabbit polyclonal antibod-ies against purifi ed Engelbreth–Holm–Swarm (EHS) tumour 14  Klinow ska et al. Copyright © 1999 by Academic Press. All rights of reproduction in any form reserved.  laminin-1 and mouse laminin E3 fragment specifi c for the laminin  1 chain (kind gifts of P. Yurchenko) w ere also raised in thislaboratory. Purifi ed IgG fractions w ere subsequently affi nity-purifi ed on EHS–Sepharose columns. Rat anti-  6 integrin mono-clonal antibody (G oH3)w as obtained from Serotec (Oxford, UK), asw as the appropriate rat IgG 2a control antibody. Pellet Preparation  Elvax pellets w ere prepared according to the methods of Silber-stein and Daniel (1982). Briefl y, 10 mg of anti-  1 integrin, anti-laminin, anti-E3, or control rabbit IgG or 1 mg of anti-  6 integrin orcontrol rat IgG 2a w as added to 20 mg bovine serum albumin(Sigma)and lyophilised. The lyophilate w as mixed w ith 250   l of a10% (w /v) Elvax (Dupont, gift of G . Silberstein) solution in dichlo-romethane and immediately frozen. The resulting pellet w as left todry overnight at   20°C and then vacuum desiccated. Antibodycontent w as calculated as w eight of incorporated material (i.e., 10or 1 mg antibody respectively) per total w eight of pellet. The largesrcinal pellet w as cut into smaller pieces (  1 mm  3 ) for implanta-tion. The typical amount of antibody implanted in these experi-ments w as 300   g/pellet piece for rabbit polyclonal or 30   g/pelletpiece for rat monoclonal antibodies. Implantation of Pellets  Age-matched 5-w eek-old virgin ICR or MF1 outbred mice w ereanaesthetised and the ventral skin w as retracted to expose bothabdominal No. 4 mammary glands. A small pocket w as made in thefat pad of one gland proximal to the lymph node (i.e., just in frontof the advancing ducts), and an antibody pellet w as inserted. Thecontralateral gland 4 either had no implant or had a control IgGpellet. The skin w as then sutured closed. The mice w ere adminis-tered analgesia and left to recover on a heated pad. BrdU Injection and Detection  To assess proliferation rates in the mammary gland, 10   l BrdUlabelling reagent (Amersham International Plc, Little Chalfont,UK) per gram of mouse body w eight w as injected intraperitoneally2 h before the glands w ere removed from the mouse and processedfor w hole-mounting, cryoembedding, or paraffi n embedding. De-tection of BrdU incorporation in paraffi n sectioned tissue w asperformed using an anti-BrdU primary antibody and an alkalinephosphatase-conjugated secondary antibody to cause a colour reac-tion w ith DAB-Ni (Amersham). Sections w ere counterstained w ith0.5% methyl green in 0.1 M sodium acetate, pH 4.0, for 15 min.Proliferation indices w ere calculated as the proportion of labelledcells in an end bud. Mammary Gland Whole-Mount  Mice w ere killed by cervical dislocation. The No. 4 glands w ereremoved and spread fl at on microscope slides, fi xed in 70%ethanol, 5% formalin, 5% glacial acetic acid, and defatted inacetone. Cell nuclei w ere stained w ith haematoxylin. The tissuew as then dehydrated sequentially in ethanol and cleared in 100%methyl salicylate. Photomicrographs w ere taken on Kodak Ekta-chrome 160T fi lm using an Olympus OM4 camera mounted on aLeica dissecting microscope. Paraffin Embedding  Number 4 glands w ere removed, cut into pieces, fi xed in 4%paraformaldehyde (1 h at 4°C), and embedded in paraffi n usingstandard techniques. Alternatively, for tissue that had already beenw hole-mounted, pieces of gland w ere cut from slides and w ashed intw o changes of xylene before being embedded in w ax. Five-micrometer sections w ere cut on a rotary microtome and mountedon uncoated glass slides. Haematoxylin and Eosin Staining (H  & E Staining)  Paraffi n sections w ere dew axed in xylene and rehydratedthrough alcohols. After a w ash in distilled w ater, slides w erestained for 1 min in Harris’s haematoxylin, “ blued” in tap w ater,and stained for 1 min in eosin. Tissue that had been w hole-mounted and stored in methyl salicylate before sectioning requiredlonger staining in eosin and less time in haematoxylin. Assessment of Apoptotic Indices  Apoptosis w as quantifi ed by examining H&E-stained paraffi nsections under an Olympus microscope. Apoptotic cells w eredefi ned as those that had dark, condensed, and lobular nuclei andw hose cytoplasm w as eosinophilic. In our experience this is amuch more reliable and reproducible indicator of apoptosis thanTUNEL labelling for this tissue. End buds w ere identifi ed and thenumber of apoptotic cells as a proportion of the total number ofcells in each end bud section w as calculated. Cryoembedding  Number 4 glands w ere removed, placed in OCT mountingmedium (Tissue-tek), and frozen. Sections (7 or 30   m)w ere placedon glass slides, fi xed in precooled methanol:acetone 1:1 at   20°C(10 min), and stored at   20°C until required. Immunocytochemistry  Immunocytochemistry w as performed as previously described(Streuli and Bissell, 1990). Anti-  1 integrin antibody w as used at 10  g/ml. Anti-laminin-1 and laminin E3 fragment antibodies w ereused at 1 and 3   g/ml, respectively. Rabbit polyclonal serum raisedin this laboratory against collagen IV w as used at 1:100 dilution.Donkey anti-rabbit FITC-conjugated secondary antibody w as usedat a concentration of 1:100 (Jackson ImmunoResearch Laborato-ries, Inc.). Nuclei w ere counterstained w ith 4   g/ml Hoechst 33258and mounted in 1 mg/ml  p  -phenylene diamine in 10% PBS, 90%glycerol, pH 8.0. Stained sections w ere examined on a Zeissepifl uorescence microscope and photographed using Tmax 400fi lm or on a Leica confocal microscope. Metabolic Labelling and Immunoprecipitation of Integrins  This w as performed as previously described (Delcommenne andStreuli, 1995). Briefl y, K-1735 M2 melanoma, CID-9, or primarymammary epithelial cells w ere radiolabelled overnight w ith[ 35 S]methionine (NEN) and lysed in 50 mM Tris, pH 8.0, 150 mMNaCl, 1% Nonidet-P40, 1 mM CaCl 2 , 1 mM MgCl 2  containing 0.45TIU/ml aprotinin, 10   M leupeptin, 0.5   M PMSF, and 1.5   M 15  1 Integrins and Mammary Morphogenesis  Copyright © 1999 by Academic Press. All rights of reproduction in any form reserved.  pepstatin A (Sigma). Aliquots of lysates containing equal TCA-precipitable counts were reacted with 10   g antibody and immunecomplexes precipitated with protein A–Sepharose (Zymed Labora-tories Inc., South San Francisco, CA). The bound complexes werewashed and analysed by SDS–PAG E under nonreducing conditionson 6.25% gels. G els were dried and autoradiographed with KodakXAR-5 fi lm at   70°C. Immunoblotting  Purifi ed laminin-1 (0.2   g)or its E3 component, EHS matrix, andpurifi ed collagen IV (Becton Dickinson Oxford, UK)were separatedunder reducing conditions on 4.0–7.5% SDS–polyacrylamide gra-dient gels, transferred to Immobilon-P membrane (Millipore LTD.,Oxford, UK), incubated with 1   g/ml affi nity-purifi ed anti-lamininor anti-E3 polyclonal antibodies and detected by enhanced chemi-luminescence using an ECL kit (Amersham). Adhesion Assays  First-passage primary mouse mammary epithelial cells, CID-9 orFSK-7 mammary epithelial cell lines (Pullan and Streuli, 1996; 4  10 4 cells/well), were incubated for 60 min in the presence of 0–400  g/ml anti-  1 integrin, anti-laminin-1, or anti-E3 IgG or 400   g/mlcontrol rabbit IgG in 96-well dishes (Maxisorb, Nunc, KampstrupRoskilde, Denmark) coated with physiological substrata (12   g/mlfi bronectin, 2   g/ml collagen IV, 12   g/ml laminin-1, and 3   g/mlvitronectin (all Sigma), or 16   g/ml collagen I and 1   g/ml EHSmatrix; Pullan and Streuli, 1996). After unattached cells werewashed away, adhered cells were fi xed, stained with crystal violet,and solubilised in 2 M guanidine hydrochloride before the absor-bance was read at 595 nm. Culture Assay for Mammary Ductal Morphogenesis  TAC-2.1 cells (Soriano  et al.,  1996), a clonally derived subpopu-lation of the TAC-2 cell line (Soriano  et al.,  1995), were cultured intissue culture fl asks (Falcon, Becton Dickinson and Co., San Jose´,CA) coated with collagen I in DME (G ibco, Basel, Switzerland)supplemented with 10% FCS (G ibco), penicillin (110 iu/ml), andstreptomycin (110   g/ml). Recombinant human HG F was a gener-ous gift of Dr. R. Schwall (G enentech Inc., South San Francisco,CA). TAC-2.1 cells were suspended at 6  10 3 cells/ml in collagengels (500   l) in 16-mm wells of four-well plates (Nunc) andincubated in 500   l medium with or without HG F and anti-  1integrin antibody. Medium and treatments were renewed every 2–3days. After 9 days, the cultures were fi xed with 2.5% glutaralde-hyde in 0.1 M cacodylate buffer, and at least three randomlyselected fi elds (measuring 2.2 mm    3.4 mm) per experimentalcondition in each of at least three separate experiments werephotographed under bright-fi eld illumination with a Nikon Dia-phot TMD inverted photomicroscope. The total length of cords ineach individual colony was measured with a Qmet 500 imageanalyser (Leyca Cambridge Ltd., Cambridge, UK). Cord length wasconsidered as 0 in (a) spherical colonies and (b) slightly elongatedstructures in which the length to diameter ratio was less than 2.Values for cord length obtained from the largest colonies are anunderestimate, since in these colonies a considerable proportion ofthe cords were out of focus and therefore could not be measured.Values are expressed as mean cord length per photographic fi eld(Soriano  et al.,  1996). The mean values for each experimentalcondition were compared to controls using Student’s unpaired  T   test. Analysis of c-Met Phosphorylation  TAC-2.1 cells were cultured on collagen I to subconfl uency,serum starved overnight, and treated with 20   g/ml HG F (R&DSystems, Abingdon, UK)and/or 50   g/ml anti-  1 integrin antibodyor control IgG . Cells were washed in ice-cold PBS containing 1 mMNa 3 VO 4 , 1 mM NaF (Sigma)and their contents were extracted withlysis buffer (1% Triton X-100, 1 mM PMSF, 50 mM Tris, pH 7.5,150 mM NaCl, 2 mM EDTA, 1 mM Na 3 VO 4 , 1 mM NaF, 10 mMleupeptin and aprotinin). After the detergent insoluble proteinswere cleared by centrifugation, equal amounts of protein (asestimated by Coomassie-stained gel electophoresis)were immuno-precipitated with anti-c-Met antibody (Santa Cruz BiotechnologyInc., Santa Cruz, CA) or control rabbit IgG followed by proteinA–Sepharose (Zymed)before separation by 7.5% SDS–PAG E. Aftertransfer to Immobilon-P membrane (Millipore), phosphorylatedproteins were revealed with the anti–phosphotyrosine antibody4G 10 (Upstate Biotechnology Inc., Lake Placid, NY) followed byenhanced chemiluminescence using an ECL kit (Amersham). Blotswere stripped according to the Amersham protocol and reprobedwith precipitating antibody. RESULTS Distribution of Extracellular Matrix Components and    6 and    1 Integrins w ithin Virgin Mammary Gland  The end buds and ducts of normal virgin mammaryglands are surrounded by a sheath of ECM containing bothlaminin and collagen IV (Figs. 1A–1D). Laminin-1 wasadjacent to the end buds as it was recognised both by theanti-laminin-1 antibody and by the anti-laminin   1 chainantibody raised against its E3 domain. Collagen IV waspresent at the end bud tips, although it was thinner in thisregion. The only previous study carefully examining thedistribution of ECM proteins around end buds, rather thanother types of mammary epithelial structures, did notdetect basement membrane components around the distalend bud region (Sonnenberg  et al.,  1986), but our investiga-tions using confocal microscopy indicate that the proteinsare indeed present.Staining for   1 integrins was seen throughout the endbud, particularly strongly at the basal surface of both theend bud epithelium and the subtending ducts (Figs. 1E–1G ).This corresponds to the cap cell layer in end buds and themyoepithelial cell layer of ducts (Williams and Daniel,1983). In addition,   1 integrins were present at sites ofcell–cell interaction within the luminal epithelium, andoccasional cells within end buds were strongly stained (Fig.1G ). There was weak immunoreactivity with stromal cells.  1 integrins have not previously been examined in mousemammary gland, but their localisation is similar to thatobserved in human breast lobules (Zutter  et al.,  1990).   6integrin was present around all epithelial cells of themammary gland and in the stroma (data not shown), as 16  Klinow ska et al. Copyright © 1999 by Academic Press. All rights of reproduction in any form reserved.  FIG.1.  Immunolocalisation of extracellular matrix molecules and   1 integrins in the virgin mammary gland. (A–D) Thick cryosections(30   m) of virgin mammary gland w ere stained for (A, B) the E3 domain of the laminin   1 chain, (C) laminin-1, and (D) collagen IV. (B) isa higher pow er image of (A). Sections w ere counterstained w ith Hoechst 33258 (blue) to detect nuclei. Projections of 10 z-sections (3   msteps)obtained by confocal microscopy are show n. Note staining for (A, B)E3 in a complete sheath of basement membrane surrounding theend bud, but not in the stroma compared (C)w ith laminin-1 show ing staining in both the basement membrane and the surrounding stroma.(D) Collagen IV is present in the end bud basement membrane as w ell as in the fi brous stroma encasing the duct. Note the lack of fi brousstroma at the end bud tip. (E–G ) Thin cryosection (7   m) of virgin mammary gland stained (E) w ith Hoescht 33258 and (F) w ith anti-  1integrin antibody. Line in (E) indicates approximate demarcation betw een end bud and subtending duct. (G ) An enlargement of the boxedarea in (F). Strong staining for   1 integrins is seen at the basal surface of the end bud epithelium and at sites of cell–cell interaction in thebasal layer. Luminal epithelial (LE), cap (CAP) and myoepithelial (ME) cells as w ell as stroma (S) are positively stained. BM, basementmembrane; FS, fi brous stroma. Bars, 50   m (A and C); 25   m (B, D, E, F); 20   m (G ). 17  1 Integrins and Mammary Morphogenesis  Copyright © 1999 by Academic Press. All rights of reproduction in any form reserved.
Related Search
Similar documents
View more
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!